Effective new strategies for the prevention and treatment of cancer will depend upon the detailed understanding of cancer evolution through its progressive stages at both the genomic and molecular levels. Since it is particularly difficult to study such changes in the human population, my laboratory has focused on the application of animal models of mammary and prostate cancer for understanding molecular processes involved in the development of these cancers and the judicious utilization of appropriate models for translational studies. The research efforts of my laboratory have focused on dissecting dynamic changes in genome organization and gene expression that occur during cancer progression in rodent models of mammary and prostate cancer and to use this information to understand how to better prevent or inhibit the oncogenic process. An important aspect of this work is to determine on a molecular level how similar various models are to the human cancers that they are thought to represent. The fundamental premise of our work is that cancer does not result from the dysregulation of a single gene, but rather from multiple, complex coordinated interactions that allow cells to grow and metastasize to a foreign habitat, ultimately killing the patient. In order to begin to decipher such complexity through a systems biology approach, my laboratory utilizes high throughput molecular techniques to amass large datasets that can be used to identify sets of genes whose collective expression correlates to genetic and biologic properties of our experimental systems. Functional testing of candidate genes is subsequently performed to validate the biologic role of particular genes. This process may be iterative.The laboratory has made significant advances in our three major areas of scientific focus:1)genomic analyses of mammary and prostate cancer development and progression, and comparison of GEM models to human breast and prostate cancer. Several microarray studies have been performed which define gene expression during normal mammary development as well as during progression of mammary cancer. A significant observation is that very few expression changes are identified between pre-invasive and invasive tumors and metastases. Genes responsive to estrogen, progesterone and prolactin have been identified in vivo. A signature which defines ER+ and ER- mammary tumors in both mouse and human has been identified. We have demonstrated that the addition of mouse array data to human data has significantly improved the class predictor for ER status in human tumors, a finding with important implication. Additionally, we have been able to classify tumors from mouse mammary cancer models in categories identified for human breast cancers, which is a major advance in understanding how these models relate to human disease. This will have high impact in determining how to use these models for pre-clinical testing. Similarly, androgen-responsive genes have been identified in two lobes of the prostate. These data are being analyzed in the context of functional changes that occur during the transition from hormone-dependent to hormone-independent tumors in both the mammary and prostate glands. Molecular differences between transgenic and chemically-induced models of prostate cancer have been identified and the molecular characteristics of these tumors are being compared to human prostate cancer. The laboratory is migrating towards more direct human cancer analyses. We have begun to analyze a set of over 200 human gastric cancer tumors including matched samples pre-and post-treatment with extensive follow-up. This will provide potentially valuable information for predicting tumor outcome based upon array signatures and response to therapy.2)in vivo pre-clinical testing of preventive and anti-angiogenic approaches of mammary cancer.